VerbOcean extractor

From WandoraWiki
Jump to: navigation, search

VerbOcean is a semantic network of verbs created by Timothy Chklovski and Patrick Pantel at the University of Southern California. VerbOcean extractor reads simple text files of VerbOcean file format and converts the file content to a Topic Map. Wandora does not include VerbOcean files and you must download such file from VerbOcean's home page or generate a VerbOcean file of your own.

VerbOcean file format

VerbOcean file format is a text file format where each line contains a quartet with two verbs and a relation, and a strenght value. Below is a small fragment of VerbOcean file:

 pursue [happens-before] abandon :: 11.262514
 abandon [opposite-of] pursue :: 10.178111
 abandon [happens-before] renew :: 10.151251
 abandon [opposite-of] renew :: 9.658213

The file may also contain comments. Comment lines start with # character.

Using VerbOcean extractor

Wandora's VerbOcean extractor starts with menu option File > Extract > Language > VerbOcean extractor. Extractor requests VerbOcean file and converts file to topic map if valid file is found.

Personal tools